This research grant combines a global and gene-specific approach to the study of DNA methylation in human breast cancer in order to identify biomarkers of breast cancer risk. Many risk factors for breast cancer are known, and can help identify cohorts of women with an increased lifetime risk of developing breast cancer, however, the majority of these women will not develop cancer and will not benefit from preventive measures. A biomarker for identifying women with a substantial risk of developing cancer in the short-term would allow appropriate selection of women and allow intervention at the time of highest risk. The ideal biomarker for stratifying risk is biologically significant in the process of malignant progression, statistically associated with cancer development, present in a significant proportion of high-risk individuals, obtainable by minimally invasive means, and reversible with known preventive interventions. We hypothesize that methylation changes in the CpG islands of tumor suppressor genes represent some of the earliest changes in the progression to malignancy in breast epithelium and can be used as a biomarker to identify women who are at short-term high-risk of developing breast cancer. Specific Aim 1: Define the patterns of DNA methylation in the CpG islands of a panel of known or putative tumor suppressor genes in breast tissues representing a spectrum of short-term breast cancer risk. Specific Aim 2: Determine whether the methylation patterns observed in the informative genes enable discrimination between cancer patients and non-cancer patients. These studies will identify genes that are biologically significant in tumorigenesis whose methylation status is statistically correlated with the presence of cancer as well as with the risk of cancer in patients with benign lesion. Our preliminary data show the feasibility of using DNA methylation patterns to stratify for risk in the esophageal cancer model, and demonstrate the capacity to undertake the methylation studies in human breast cancer specimens. The long-term goals of this study are to identify a panel of genes whose methylation status can be used as a biomarker of short-term breast cancer risk. Once validated, this model can then be applied to high-risk women entering chemoprevention trials to determine whether the methylation changes are reversible by known chemopreventive modalities. These future studies will determine whether DNA methylation changes can be used a) to stratify women at risk for breast cancer and b) as a surrogate endpoint in chemoprevention trials.